De Novo Pyrimidine BiosynthesisConnects Cell Integrity to AmphotericinB Susceptibility in Cryptococcusneoformans

Dithi Banerjee,a Timothy C. Umland,b* John C. Panepintoa

Department of Microbiology and Immunology, Witebsky Center for Microbial Pathogenesis and Immunology,University at Buffalo, The State University of New York, Buffalo, New York, USAa; Department of StructuralBiology, Hauptman Woodward Medical Research Institute, University at Buffalo, The State University of NewYork, Buffalo, New York, USAb

ABSTRACT The use of amphotericin B (AmB) in conjunction with 5-fluorocytosine(5-FC) is known to be the optimal therapy for treating cryptococcosis, but the mech-anism by which 5-FC synergizes with AmB is unknown. In this study, we generated aCryptococcus neoformans ura1Δ mutant lacking dihydroorotate dehydrogenase(DHODH), which demonstrated temperature-sensitive growth due to a defect incell integrity and sensitivity to cell wall-damaging agents. In addition, sensitivityto AmB was greatly increased. Inclusion of uracil or uridine in the medium didnot suppress the cell wall or AmB phenotype, whereas complementation withthe wild-type URA1 gene complemented the mutant phenotype. As a measure ofmembrane accessibility, we assayed the rate of association of the lipid-bindingdye 3,3=-dihexyloxacarbocyanine iodide (DiOC6) and saw more rapid associationin the ura1Δ mutant. We likewise saw an increased rate of DiOC6 association inother AmB-sensitive mutants, including a ura� spontaneous URA5 mutant made by5-fluoroorotic acid (5-FOA) selection and a bck1Δ mutant defective in cell integritysignaling. Similar results were also obtained by using a specific plasma membrane-binding CellMask live stain, with cell integrity mutants that exhibited increased andfaster association of the dye with the membrane. Chitin synthase mutants (chs5Δand chs6Δ) that lack any reported cell wall defects, in turn, demonstrate neither anyincreased susceptibility to AmB nor a greater accessibility to either of the dyes. Fi-nally, perturbation of the cell wall of the wild type by treatment with the �-1,6-glucan synthase inhibitor caspofungin was synergistic with AmB in vitro.

IMPORTANCE Synergy between AmB and nucleotide biosynthetic pathways hasbeen documented, but the mechanism of this interaction has not been delineated.Results from this study suggest a correlation between uridine nucleotide biosynthe-sis and cell integrity likely mediated through the pool of nucleotide-sugar conju-gates, which are precursor molecules for both capsule and cell wall of C. neofor-mans. Thus, we propose a mechanism by which structural defects in the cell wallresulting from perturbation of pyrimidine biosynthesis allow faster and increasedpenetration of AmB molecules into the cell membrane. Overall, our work demon-strates that impairment of pyrimidine biosynthesis in C. neoformans could be a po-tential target for antifungal therapy, either alone or in combination with AmB.

KEYWORDS amphotericin B, cell wall integrity, pyrimidine biosynthesis

The pathogenic fungus Cryptococcus neoformans is associated with fatal meningitis,most frequently infecting patients with AIDS or other immune defects resulting

from transplantation or chemotherapy (1). The recommended antifungal regimen for

Received 7 July 2016 Accepted 24 October2016 Published 16 November 2016

Citation Banerjee D, Umland TC, Panepinto JC.2016. De novo pyrimidine biosynthesisconnects cell integrity to amphotericin Bsusceptibility in Cryptococcus neoformans.mSphere 1(6):e00191-16. doi:10.1128/mSphere.00191-16.

Editor J. Andrew Alspaugh, Duke UniversityMedical Center

Copyright © 2016 Banerjee et al. This is anopen-access article distributed under the termsof the Creative Commons Attribution 4.0International license.

Address correspondence to John C. Panepinto,jcp25@buffalo.edu.

�Present address: Timothy C. Umland,QuaDPharma, LLC, Clarence, New York, USA.

RESEARCH ARTICLETherapeutics and Prevention

crossmark

Volume 1 Issue 6 e00191-16 msphere.asm.org 1

on February 24, 2021 by guest

http://msphere.asm

.org/D

ownloaded from

the treatment of cryptococcosis is 2 weeks of amphotericin B (AmB) in combinationwith 5-fluorocytosine (5-FC), a pyrimidine analogue, followed by fluconazole mono-therapy (2–4). Studies of fungal burden in cerebrospinal fluid (CSF) in patients withcryptococcosis demonstrate a clear enhancement of the fungicidal activity of AmB incombination with 5-FC compared to AmB alone or AmB in combination with flucona-zole (5, 6). In vitro work demonstrates synergy between AmB and 5-FC as well asbetween AmB and mycophenolic acid, an inhibitor of the IMP dehydrogenase in the denovo purine biosynthesis pathway (7, 8). The exact mechanism by which AmB and 5-FCsynergize is still unknown. The nucleotide biosynthesis pathway is therefore an inter-esting avenue for further research, with emphasis on its role in potentiation of AmBefficacy.

Dihydroorotate dehydrogenase (DHODH) is a component of the pyrimidine biosyn-thesis pathway. Two families of DHODH enzymes exist in evolution, family 1 and family2 (9). Family 1 enzymes are cytosolic and utilize a cysteine as the catalytic residue in theactive site (10, 11). Family 2 enzymes, in contrast, are localized to the mitochondrialmembrane and contain a serine as the catalytic residue in the active site (10). DHODHenzymes are known to be targetable by small molecules but exhibit enough variationthrough evolution that species specificity can be achieved in molecule design (12).Brequinar is a known inhibitor of the human DHODH and a family 2 enzyme and is usedin immunosuppressive therapy for transplant rejection (13, 14). The malaria parasitealso expresses a family 2 enzyme, but the malarial DHODH is not sensitive to brequinar(12, 15), indicating the species specificity of this enzyme. Impairment of de novopyrimidine biosynthesis by perturbing other enzymes in the pathway impairs cellintegrity in C. neoformans and prevents capsule biosynthesis in the absence of exog-enous uracil (16, 17).

In this study, we used a Cryptococcus neoformans var. grubii DHODH mutant, a ura1Δstrain, as a tool to investigate the mechanism by which perturbations in pyrimidinebiosynthesis lead to potentiation of AmB fungicidal activity. Deletion of URA1 resultedin cell integrity defects, temperature sensitivity, mitochondrial defects, replication stresssensitivity, and hypersensitivity to AmB. Rates of association of the membrane-bindingdyes 3,3=-dihexyloxacarbocyanine iodide (DiOC6) and CellMask green were increased inthe ura1Δ mutant, suggesting that loss of de novo pyrimidine biosynthesis increasesmembrane accessibility. DiOC6 and CellMask green association rates were also in-creased in a C. neoformans bck1Δ mutant, lacking the mitogen-activated protein (MAP)kinase kinase kinase (MAPKKK) in the cell integrity MAP kinase pathway, which was alsofound to be AmB sensitive. Two chitin synthase mutants, however, did not displayincreased membrane accessibility to the lipophilic dyes and exhibited wild-type AmBsensitivity. Our results have led us to propose a mechanism by which perturbation ofde novo pyrimidine biosynthesis impairs cell integrity, thereby allowing increasedaccess of AmB to the cell membrane, resulting in increased susceptibility.

RESULTSC. neoformans encodes a single family 2 DHODH required for wild-type growth.Cryptococcus neoformans dihydroorotate dehydrogenase belongs to the family 2DHODHs (Fig. 1), which include the enzymes largely from mammals, higher eukaryotes,and prokaryotes as well (10, 12, 18–20). Family 2 DHODH enzymes from mammals andhigher eukaryotes typically possess N-terminal extensions that facilitate their mitochon-drial localization and association with mitochondrial membrane. On the other hand,this N-terminal extension tends to be absent in prokaryotic family 2 DHODH, and theprotein is bound to the cytoplasmic membrane, in contrast to family 1 enzymes, whichare cytosolic (9, 10). Further traits distinguishing DHODH families are their electronacceptor cofactors. Family 1 DHODH uses either a fumarate or NAD� cofactor, whereasfamily 2 DHODH employs ubiquinone binding in a hydrophobic channel near the Nterminus (21). Finally, serine replaces cysteine of the family 1 DHODH enzymes as thecatalytic residue and is conserved among all the family 2 DHODH enzymes (10).

Banerjee et al.

Volume 1 Issue 6 e00191-16 msphere.asm.org 2

on February 24, 2021 by guest

http://msphere.asm

.org/D

ownloaded from

FIG 1 C. neoformans encodes a single family 2 DHODH required for wild-type growth. (A) Multiple sequence alignment of C. neoformans Ura1 proteinsequence (top line) with other eukaryotic family 2 DHODHs including (from top to bottom) Ustilago maydis (DQ869012), Candida albicans (AY230865),Schizosaccharomyces pombe (NM_001018748), human (NM001361), and Plasmodium falciparum (AB070244), and the family 1 enzyme from S. cerevisiae(P28272) (bottom line), with only the respective N-terminal regions shown. The first 25 amino acids are important for mitochondrial import and membrane

(Continued)

Cell Integrity Defects Synergize with AmB

Volume 1 Issue 6 e00191-16 msphere.asm.org 3

on February 24, 2021 by guest

http://msphere.asm

.org/D

ownloaded from

Studies from humans, Plasmodium falciparum, and other fungal species, includingCandida albicans, Schizosaccharomyces pombe, and Ustilago maydis, demonstrate thestructure of DHODH enzymes to be comprised of two domains, an N-terminal alpha-helical domain and a large C-terminal domain connected by an extended loop (9, 15,18, 22). The predicted annotation of the C. neoformans var. grubii DHODH(CNAG_02794) appears to lack the N-terminal extension present in all other eukaryoticfamily 2 DHODH enzymes analyzed (Fig. 1A), with an N terminus similar in length tothat of prokaryotic family 2 DHODH (data not shown). Our own 5= rapid amplificationof cDNA ends (RACE) data confirm the annotated transcriptional start site. Superimpo-sition of the C. neoformans sequence on the structure of human DHODH revealed anabsence of a portion of this N-terminal helical region in C. neoformans. Despite thistruncation, the cryptococcal N-terminal sequence bears a predicted mitochondrialtargeting signal peptide (iPSORT prediction tool) and contains the protruding hydro-phobic patch with identical positively charged residues (His14, Arg15, Leu16, Arg19,Arg28, and Arg30) (10). The remaining residues important for cofactor binding, sub-strate binding, and catalytic activity are completely conserved in the C. neoformansDHODH sequence (Fig. 1A), including the catalytic serine that is a hallmark of family 2DHODH enzymes in contrast to the catalytic cysteine in family 1 enzymes (10).

We generated a deletion mutant in the C. neoformans var. grubii strain H99, a fullyvirulent and melanizing derivative of H99 (23). We then introduced the wild-type URA1gene in trans to generate a complemented mutant. Deletion of URA1 resulted in uracilauxotrophy that was only partially remediated by the addition of uracil and/or uridineto the medium (Fig. 1B). Reintroduction of the wild-type URA1 gene completelyrestored uracil prototrophy and wild-type growth. This result suggests that pyrimidinesalvage is insufficient for wild-type growth of C. neoformans.

The production of the polysaccharide capsule draws from nucleotide pools, as eachcarbohydrate monomer is added from a nucleotide-sugar conjugate (24, 25). De novopyrimidine biosynthesis is required for capsule production in the absence of exogenousuracil. Consistent with these reports, the ura1Δ mutant was defective for capsuleproduction both at 37°C (data not shown) and at 30°C (Fig. 1C). The addition of uracilor uridine did not result in capsule synthesis in the acapsular ura1Δ strain (Fig. 1C).Capsule induction was observed after 24, 48, and 72 h of incubation, and the meancapsule size of each strain was tabulated as a bar graph (Fig. 1D), which indicatedthat capsule induction took place between 24 and 48 h of incubation and that therewas no change in capsule size from 48 to 72 h. The fungicidal effect of AmB isfast-acting, as killing is seen within the first 6 to 12 h of exposure by time-kill assays(26–28). Thus, the capsule is likely not contributing significantly to the susceptibility ofC. neoformans to AmB.

DHODH is required for wild-type cell integrity. The cell wall of C. neoformansis made up of five polysaccharides, including �-1,6-glucan, �-1,3-glucan, �-1,3-glucan,chitin (�-1,4-GlcNAc), and chitosan (deacetylated �-1,4-GlcNAc). Because each mono-mer in these large polysaccharide polymers is derived from a UDP-sugar conjugate, itfollows that defects in pyrimidine biosynthesis should result in impaired cell wallsynthesis (24). To assess the cell integrity of the C. neoformans ura1Δ mutant, weperformed spot plate assays in the presence of known cell integrity stressors, includingthe detergent SDS and the �-1,3-glucan synthase inhibitor caspofungin. As demon-strated in Fig. 2B, the ura1Δ mutant exhibited sensitivity to the stressors tested, and thereintroduction of the URA1 gene restored wild-type sensitivity.

Cell integrity is inextricably linked to temperature sensitivity, and so we comparedthe ability of the ura1Δ mutant to grow on rich medium at 30°C, 37°C, and 39°C with

Figure Legend Continuedanchorage. The boxes highlight the hydrophobic residues that make up the hydrophobic patches in the family 2 DHODHs. (B) Growth of wild-type, ura1�mutant, and complemented strains in rich medium (YPD) or minimal medium with and without uracil supplementation. (C) India ink staining ofC. neoformans capsule under inducing conditions in the wild type, ura1� mutant, complemented mutant, and ura1� strain supplemented with uridineat 24, 48, and 72 h. (D) Mean capsule sizes of wild type, ura1� mutant, ura1� mutant supplemented with uridine, and complement strain.

Banerjee et al.

Volume 1 Issue 6 e00191-16 msphere.asm.org 4

on February 24, 2021 by guest

http://msphere.asm

.org/D

ownloaded from

those of the wild type and complemented mutant (Fig. 2A). As expected, the ura1Δmutant was sensitive to growth at 37°C and was unable to grow at 39°C. This result isconsistent with known attenuation in virulence of C. neoformans pyrimidine biosyn-thesis mutants. To investigate the link between the temperature sensitivity and cellintegrity defects of the ura1Δ mutant, we assayed the ability of the osmostabilantsorbitol to rescue temperature-sensitive growth in the ura1Δ mutant. Figure 2C dem-onstrates that sorbitol was unable to rescue the temperature-sensitive growth pheno-type of the ura1Δ mutant, suggesting that additional factors beyond cell integritycontribute to its temperature sensitivity.

Mitochondrial function and replication stress resistance are impaired in aura1� mutant. It is known that mitochondrial function is induced in C. neoformans inresponse to host temperature (29). DHODH is a mitochondrial enzyme, and depletionof DHODH in mammalian cells induces mitochondrial dysfunction (30, 31). To deter-mine if the ura1Δ mutant exhibited mitochondrial defects, we assayed sensitivity tocompounds that perturb mitochondrial function, including the respiratory chain inhib-itor antimycin A and oxidative stress inducer H2O2. Figure 3A demonstrates that theura1Δ mutant was sensitive to mitochondrial perturbation. To determine if this phe-notype was unique to the ura1Δ mutant or was shared with other de novo pyrimidinebiosynthesis mutants, we assayed the sensitivity of a spontaneous ura5 mutant and itscomplement in parallel. Figure 3A demonstrates a shared sensitivity to mitochondrial

FIG 2 DHODH is required for wild-type cell integrity. Spot plate analysis of WT, ura1� mutant, andcomplemented strains on rich medium at the indicated temperatures (A), in the presence of cell wallstressors (B), and in the presence of added osmostabilant (C). Caspo, caspofungin; Sorb, sorbitol.

FIG 3 Mitochondrial function and replication stress resistance are impaired in a ura1� mutant. Spotplate analysis of WT, ura1� mutant and C-URA1, and ura5 mutant and C-URA5 strains with mito-chondrial stressors (A) and replication stress inducers (B). AntA, antimycin A.

Cell Integrity Defects Synergize with AmB

Volume 1 Issue 6 e00191-16 msphere.asm.org 5

on February 24, 2021 by guest

http://msphere.asm

.org/D

ownloaded from

perturbation, suggesting that the defects stem from lack of flux through the pathwayrather than just the physical absence of the DHODH enzyme.

The homeostasis of nucleotide pools is important for genome integrity in eukaryoticcells. Because pyrimidine salvage is insufficient to fully rescue ura1Δ mutant pheno-types, we went on to determine if the ura1Δ mutant was sensitive to replication stress.Figure 3B demonstrates that both a ura1Δ and a ura5 spontaneous mutant are deficientin their response to replication stress, including alkylation by treatment with methylmethanesulfonate (MMS), inhibition of ribonucleotide reductase by hydroxyurea (HU),or induction of DNA cross-links by exposure to UV light. Each phenotype was restoredcompletely by reintroduction of the wild-type gene.

DHODH is required for virulence in a Galleria mellonella infection model.The larvae of Galleria mellonella are a powerful surrogate model for assessing crypto-coccal pathogenicity, especially as it relates to temperature adaptation (32). To inves-tigate the contribution of ura1Δ mutant temperature sensitivity to pathogenesis, wecompared the abilities of the wild type, the ura1Δ mutant, and its complement to causemortality in larvae incubated at either 30°C or 37°C. Figure 4 demonstrates a rapidkilling of G. mellonella larvae by the wild-type H99 strain and URA1-complementedstrains in comparison to the ura1Δ mutant, indicating the requirement of DHODH forvirulence in an infection model of cryptococcosis. Wild-type and complemented strainskilled all the larvae by day 4 at both 37°C (Fig. 4B; P � 0.05) and 30°C (Fig. 4A; P � 0.05),whereas 100% survival was observed in larvae infected with the knockout strains atboth temperatures. These data demonstrated that in addition to temperature sensitiv-ity, pyrimidine pool depletion also plays a role in virulence, similarly to results obtainedwith a ura4Δ mutant (16). No death was observed in any of the mock-injected controlsat either 30°C or 37°C until 5 days postinoculation.

Loss of DHODH sensitizes C. neoformans to amphotericin B. Perturbations inde novo pyrimidine biosynthesis by mutation or treatment with nucleoside analoguespotentiate the fungicidal activity of AmB. To investigate the AmB sensitivity of theura1Δ mutant, we performed both Etest and broth microdilution MIC assays. As

FIG 4 DHODH is required for virulence in a G. mellonella infection model. G. mellonella larvaewere inoculated with WT, ura1�, and C-URA1 strains and monitored for survival at both 30°C (A)and 37°C (B).

Banerjee et al.

Volume 1 Issue 6 e00191-16 msphere.asm.org 6

on February 24, 2021 by guest

http://msphere.asm

.org/D

ownloaded from

expected, the ura1Δ mutant was found to be sensitive to AmB with a MIC of0.004 �g/ml (Fig. 5A), similar to that of a spontaneous ura5 mutant selected on5-fluoroorotic acid (5-FOA) in both assays (17), in comparison to 0.064 �g/ml and0.0094 �g/ml, the MICs of the wild type and the C-URA1 mutant, respectively. Etest inthe presence of exogenous uracil and uridine also showed lowered MIC levels(0.004 �g/ml) in the ura1Δ mutant (Fig. 5A), which suggests that pyrimidine salvage isinsufficient to compensate for the higher sensitivity to AmB in the absence of de novosynthesis.

The mechanism by which perturbation of pyrimidine biosynthesis potentiates AmBactivity is unknown. A simple hypothesis is that the cell integrity defects induced byinhibition or mutation of de novo synthesis allow increased access of small moleculesto the cell membrane. To test this hypothesis, we compared the rates of association ofthe membrane-binding dye DiOC6 with the wild-type, ura1Δ mutant, and comple-mented strains (33). As demonstrated in Fig. 5B, the fluorescence intensity of the ura1Δmutant increased more rapidly and to a higher equilibrium than that of either the wildtype or complemented mutant, suggesting that the cell wall of the ura1Δ mutantallowed for more rapid penetration and membrane binding of DiOC6 than of either thewild type or the complemented mutant. We assessed if this phenotype was shared withthe spontaneous ura5 mutant made by 5-FOA selection and found an increase in therate of DiOC6 association similar to that of the ura1Δ mutant (Fig. 5C). To determine ifthis increased dye association was indeed a result of faster and easier permeability tolipophilic small molecules, we performed a similar experiment with CellMask greenstain, another membrane-binding dye that is specific to the plasma membrane (34).Results demonstrate higher relative fluorescence units (RFU) in the ura1Δ mutant overa shorter period of incubation, which suggests a faster accessibility of the dye to theplasma membrane in the mutant compared to either the wild type or complement(Fig. 5D).

Cell integrity defects correlate with AmB sensitivity. If the mechanism of AmBpotentiation in the ura1Δ mutant is increased access to the membrane, we wouldexpect that other mutants with cell integrity defects would exhibit sensitivity to AmB.We selected a panel of cell integrity mutants from the Lodge Lab collection obtainedthrough the Fungal Genetic Stock Center (http://www.fgsc.net). The selected mutantswere bck1� (lacking the MAPKKK in the cell integrity MAP kinase cascade) and chs5Δand chs6� (mutants of chitin synthase genes CHS5 and CHS6, respectively) strains. Weperformed Etests and broth microdilution assays to determine the MIC of AmB for eachof the strains. Results revealed an increased sensitivity of the bck1� mutant (MIC,0.004 �g/ml), whereas the chs5� and chs6� mutants demonstrated wild-type MIClevels of 0.047 �g/ml (Fig. 6A). The result from the bck1� strain suggested that thisincreased sensitivity to AmB could be due to the defect in cell integrity, consistent withour hypothesis (35). We know that there are eight CHS genes and they are redundantin function, except for CHS3 (36). Previous studies have demonstrated notable pheno-types in only the chs3� strain and mild phenotypes in chs6� knockout strains. Webelieve that chs5� and chs6� mutants do not have significant cell integrity impairmentand thus exhibit normal sensitivity to AmB. To verify this hypothesis, we also performedDiOC6 and CellMask green staining with these strains to investigate their membraneaccessibility to the dyes. Results in Fig. 6B and C demonstrated faster and greaterpermeativity of DiOC6 and CellMask dyes, respectively, in the bck1� mutant andwild-type dye binding capacity in the chs5� and chs6� mutants that correlated with ourfindings from the AmB MIC assays. Taken together, our data suggest that a defect in cellintegrity potentiates the antifungal efficacy of AmB.

Perturbation of cell wall synthesis also potentiates AmB fungicidal activity.Finally, because cell integrity is linked to AmB susceptibility, we expected that inhibi-tion of cell wall synthesis by the �-1,3-glucan synthase inhibitor caspofungin wouldlead to potentiation of AmB activity. To test this, we quantified the interaction betweencaspofungin and AmB by checkerboard MIC analysis (37, 38). Results demonstrated inFig. 7 indicate that the mean MIC of AmB alone from three independent experiments

Cell Integrity Defects Synergize with AmB

Volume 1 Issue 6 e00191-16 msphere.asm.org 7

on February 24, 2021 by guest

http://msphere.asm

.org/D

ownloaded from

FIG 5 Loss of DHODH sensitizes C. neoformans to amphotericin B. (A) Etest analysis of AmBsensitivity in the wild type, ura1� mutant, and complemented strain on AM3 medium alone or withthe ura1� mutant in the presence of uracil or uridine. (B and C) Comparison of the rates ofassociation of DiOC6 stain between the ura1� mutant (B) or the ura5 mutant (C) and the wild typeor the respective complemented strains. (D) Comparison of the rates of association of CellMask greenlive stain between the ura1� mutant, wild type, and complemented mutant.

Banerjee et al.

Volume 1 Issue 6 e00191-16 msphere.asm.org 8

on February 24, 2021 by guest

http://msphere.asm

.org/D

ownloaded from

FIG 6 Cell integrity defects correlate with AmB sensitivity. (A) Etest analysis of AmB sensitivity in thewild type, chs5� mutant, chs6� mutant, and bck1� mutant. (B and C) Comparison of the rates ofassociation of DiOC6 stain (B) and CellMask green live stain (C) between the wild type, chs5� mutant,chs6� mutant, and bck1� mutant strains.

Cell Integrity Defects Synergize with AmB

Volume 1 Issue 6 e00191-16 msphere.asm.org 9

on February 24, 2021 by guest

http://msphere.asm

.org/D

ownloaded from

was 0.06 �g/ml and that of caspofungin alone was 40 �g/ml. Mean MIC was signifi-cantly lowered by approximately 3-fold to 0.0075 �g/ml in AmB and by 4-fold incaspofungin to 10 �g/ml. The fractional inhibitory concentration (FIC) index calculatedfrom the MIC values to determine the drug interaction is reported in Table 1. An FICvalue of �0.5 indicates synergy, a value ranging between 0.5 and 1.0 indicates additiveinteraction, any value between 1.0 and 4.0 is suggestive of indifference, and a value of�4 indicates antagonism (39). We found that a combination of caspofungin and AmBwas synergistic in action with an FIC index of 0.375, and this result further suggests thata defect in cell wall synthesis potentiates the antifungal activity of AmB.

DISCUSSION

Polyene antifungals exert their activity through interactions with ergosterol in thefungal cell membrane (40). The functional consequence of this interaction is thought tobe pore formation and cell disruption, although other modes of action have beenproposed. The potentiation of AmB activity by 5-FC is known, and inclusion of 5-FC incombination with AmB is the recommended initial therapy for treatment of cryptococ-cosis. The use of AmB in underresourced areas of the world is hampered by its toxicity(41). Defining the underlying mechanism of this synergy could inform future drugdesign to improve the activity of AmB and possibly reduce the required dose to achieveeffective antifungal activity. Reduced dosage could improve safety and widen avail-ability for use of this potent antifungal drug.

DHODH is a known druggable target. Brequinar and leflunomide target the mam-malian DHODH (14, 21). Compounds that specifically target the malarial DHODH are

FIG 7 Perturbation of cell wall synthesis also potentiates AmB fungicidal activity. Checkerboardanalysis of the activity of AmB and caspofungin in combination against wild-type C. neoformans.

TABLE 1 FIC index of AmB and caspofungin drug interaction in wild-type C. neoformans

Drug or parameter

Concn (�g/ml)

Replicate 1 Replicate 2 Replicate 3

AmB alone (MICA) 0.06 0.06 0.06AmB in combination (MICA=) 0.0075 0.0075 0.0075Caspofungin alone (MICC) 40 40 40Caspofungin in combination (MICC=) 10 10 10FIC index � MICA=/MICA � MICC=/MICC 0.375 0.375 0.375Interaction Synergy (�0.5) Synergy (�0.5) Synergy (�0.5)

Banerjee et al.

Volume 1 Issue 6 e00191-16 msphere.asm.org 10

on February 24, 2021 by guest

http://msphere.asm

.org/D

ownloaded from

currently in clinical trials (42). The C. neoformans var. grubii DHODH is distinct fromother eukaryotic family 2 DHODH enzymes in that the hydrophobic channel producedby the N-terminal alpha-helices is truncated. This makes C. neoformans DHODH moreinteresting since this trait is shared by prokaryotic family 2 DHODHs. Because thisregion corresponds to the binding region of known inhibitors, it is unlikely that theC. neoformans enzyme will be effectively inhibited by molecules that target otherfungal family 2 enzymes. C. neoformans is insensitive to brequinar, an inhibitor of thehuman DHODH (data not shown). This is unsurprising, as the closely related basidi-omycete Ustilago maydis, which encodes a family 2 DHODH with the conservedN-terminal extension, is also resistant to brequinar (18). The resistance in these fungi islikely not due to permeability of the drug, as replacement of the U. maydis DHODH withthe human enzyme results in brequinar sensitivity.

The utility of inhibiting de novo pyrimidine biosynthesis in organisms with intactsalvage pathways has been questioned. However, studies from C. neoformans demon-strate that de novo synthesis of pyrimidines is required for virulence in animal modelsof infection (16, 43). C. neoformans is unique among the pathogenic fungi in that itgenerates a large polysaccharide capsule that, in addition to the cell wall polysaccha-ride, utilizes nucleotide sugar monomers in its production. Thus, the salvage pathwayof nucleotide biosynthesis alone may not be sufficient to support the construction ofthese polymers in addition to maintenance of genome integrity and RNA synthesis inthe absence of de novo synthesis. Our time course data as well as the investigation ofcapsule regulators suggest that there is no correlation between capsule production andAmB susceptibility. Both acapsular and hypercapsular regulatory mutants display variedpatterns of susceptibility to AmB (35, 44–46).

Inhibition of de novo pyrimidine biosynthesis does impair cell wall synthesis andmitochondrial function, both of which are required for adaptation to host temperature.Increased temperature induces mitochondrial function and results in oxidative damageof important proteins. We have shown in Fig. 3A that loss of DHODH sensitizesC. neoformans to the mitochondrial stressor antimycin A and the oxidative stressorhydrogen peroxide, suggesting that this protein is required to protect the cells fromoxidative damage. This leads us to believe that in the absence of DHODH, oxidativestress, in addition to cell integrity defects, contributes to the temperature sensitivity ofthe ura1Δ mutant. Thus, an inhibitor of de novo pyrimidine biosynthesis would impairtwo additional distinct pathways required for pathogenesis. Our data using the Galleriamodel of infection demonstrate that temperature sensitivity alone cannot account forthe decrease in virulence, as equivalent defects in virulence were seen in the larvaewhen incubated at 30°C and at 37°C. It follows that inhibitors of de novo nucleotidebiosynthesis should not be discounted in the design of novel anticryptococcal agents.

The data presented in this study support a model in which inhibition of pyrimidinebiosynthesis causes defects in the cell wall of C. neoformans that, in turn, increaseaccess of small molecules such as AmB to the cell membrane. Thus, small-moleculeinhibitors of nucleotide biosynthesis, cell wall synthesis, and cell integrity signalingpathways should uniformly result in increased sensitivity to AmB. In addition, pharma-cological agents that are modestly effective in monotherapy, such as caspofungin,could be highly effective in the context of combination therapy.

MATERIALS AND METHODSStrains. C. neoformans strain H99 (serotype A), a ura5 spontaneous mutant, and the C-URA5 complementstrain were described previously (17). The bck1Δ (35) and chs5Δ and chs6Δ (36) strains were obtainedfrom the deletion collection of Jennifer Lodge through the Fungal Genetic Stock Center at Kansas StateUniversity.

ura1Δ knockout and C-URA1 complement strains were produced as follows. A C. neoformans ura1Δstrain was constructed as described previously (47). The construct was verified by sequencing, whichshowed retention of 610 bp of the URA1 gene as the 5= flanking sequence. This partial knockoutconstruct was biolistically transformed in wild-type H99 (48), and colonies were screened for uracilauxotrophy by simultaneous serial passage in yeast extract-peptone-dextrose (YPD) broth and minimalyeast nitrogen base (YNB) medium supplemented with 2% dextrose (Dex). The clones that grew in YPDbut failed to grow in the absence of exogenous uracil in YNB-2% Dex were verified by Northern and

Cell Integrity Defects Synergize with AmB

Volume 1 Issue 6 e00191-16 msphere.asm.org 11

on February 24, 2021 by guest

http://msphere.asm

.org/D

ownloaded from

Southern blot analysis. Because pyrimidine-rich tracts that flank the URA1 gene precluded use of PCR, acomplementation construct was isolated by screening a size-selected genomic library. Wild-type H99genomic DNA was first digested with EcoRI and BamHI (EcoRI and BamHI restriction sites present at 5=and 3= flanking regions of the URA1 gene) to generate a product of 3,547 bp which encompasses theURA1 gene. This 3.5-kb piece was ligated into pBluescript plasmid, and colonies were selected byblue-white screening followed by colony hybridization probing with the URA1 coding region. Theconstruct was then transformed into the ura1Δ knockout strain by electroporation (49). Transformantswere selected on minimal medium lacking uracil, confirming uracil prototrophy. Northern blot analysisconfirmed expression of wild-type DHODH. Primer sets used in this study are available from the authorsupon request.

Media and reagents. C. neoformans strains were maintained in yeast extract-peptone-dextrose (YPD;BD Difco) broth supplemented with 80% glycerol (Fisher Chemicals) as frozen stocks and freshly streakedon YPD agar (BD Difco) before experiments. Spot plate assays were performed on YPD agar alone orsupplemented with different reagents. Uracil auxotrophy was detected on yeast extract nitrogen basewith ammonium sulfate (BD Difco) agar alone or supplemented with 2% dextrose (YNB-2% Dex) or uracil(Sigma-Aldrich), uridine (Sigma-Aldrich), or both. RPMI 1640 (Gibco) was used for capsule induction andstaining assays. Antibiotic medium 3 (AM3) (Difco, MD, USA) buffered to pH 7.0 with 10 mM phosphatewas used for the Etest, and AM3 supplemented with 2% Dex was used for microbroth MIC detectionassays and checkerboard assays.

Protein structure modeling. The three-dimensional structure of C. neoformans DHODH was pre-dicted via homology modeling using the Swiss-Model server (50). A crystal structure of human DHODH(PDB 2PRL; chain A, residues 30 to 396) was used as the structural template, possessing 43.7% sequenceidentity to the C. neoformans homologue.

Spot plate assays. C. neoformans wild-type, mutant, and complemented strains were grown asovernight cultures, washed twice, resuspended, and diluted to an optical density at 600 nm (OD600) of1.00 in sterile distilled water. Tenfold dilutions of wild-type H99, ura1Δ mutant, C-URA1 complementstrain, ura5 spontaneous mutant, and C-URA5 complement cells were plated onto YPD agar plates aloneor supplemented with 0.02% sodium dodecyl sulfate (SDS; Invitrogen) and 20 �g/ml caspofungindiacetate (Sigma Aldrich) for investigating cell integrity stress, 5 �g/ml antimycin A (Sigma-Aldrich) and2 mM hydrogen peroxide (H2O2; Fisher) for mitochondrial stress, or 100 mM hydroxyurea (HU; Sigma-Aldrich) and 0.025% methyl methanesulfonate (MMS; Sigma-Aldrich) for replication stress and incubatedat 30°C for 48 h. Additionally, YPD agar plates were spotted with the control and test strains, irradiatedwith 20 mJ UV, and incubated at 30°C to investigate replication stress. Inoculated YPD agar plates withand without sorbitol (Fisher) were also incubated at 30°C, 37°C, and 39°C for determining effects withtemperature stress.

Capsule detection assay. Wild-type, ura1Δ mutant, and C-URA1 complement strains were grown tomid-log phase in YPD. Cells were then harvested, washed twice, resuspended in sterile distilled water,and adjusted to an OD600 of 1.0. Two hundred microliters of this cell suspension was inoculated in 5 ml1� RPMI alone or supplemented with uracil or uridine in a tissue culture plate for capsule induction andincubated at 30°C in 5% CO2 for 48 to 72 h. India ink preparations were made as mentioned previously(17) at 24, 48, and 72 h of incubation, and capsule was observed using LAS-AF software (Leica).

Membrane staining assay using DiOC6 and CellMask live stain. Cells (105) from overnight culturesof wild-type (WT), mutant, and complement strains were washed and resuspended in phosphate-buffered saline (PBS; Gibco), mixed with 200 nM DiOC6 (Life Technologies, Inc.) stain or 100� CellMaskstain (Thermo Fisher) for membrane staining. Cells were incubated in 96-well black-bottomed Greinerplates, and readings for relative fluorescence units (RFU) over 30 min of incubation were taken by aspectrophotometric plate reader (SpectraMax M5; Molecular Devices, CA) with excitation at 482 nm andemission at 504 nm for DiOC6 and excitation at 485 nm and emission at 535 nm for CellMask green livestain.

MIC detection by Etest. Cells (0.5 � 107) of wild-type, ura1Δ mutant, and C-URA1 complementstrains were plated onto AM3 alone or supplemented with 20 �M uracil or 20 �M uridine, and bck1Δ,chs5Δ, and chs6Δ strains were grown on AM3 alone as a lawn culture with a sterile cotton swab. Plateswere dried completely, after which amphotericin B Etest strips (BioMérieux) were placed in the center ofthe plates. Following 48 to 72 h of incubation at 30°C, the MIC was determined for each of the strainsby a standard method (17).

MIC detection by broth microdilution method. Suspensions of wild-type, ura1Δ mutant, andC-URA1 complement colonies were made in sterile distilled water, washed three times, and diluted toyield a final concentration of 1 � 107 cells. Working solutions of amphotericin B (Cellgro; Mediatech, Inc.)were made in AM3 at double the concentration that was used in the assay. One hundred microliters ofincreasing concentrations of the drug was dispensed in the 96-well microtiter plate followed by 100 �lof the inoculum to reach a final volume of 200 �l. Final concentrations of AmB ranged from 0.0009 �g/mlto 0.5 �g/ml in the assay. A well containing 100 �l each of assay medium and inoculum served as growthcontrol (GC), whereas a single well containing 200 �l of the assay medium alone served as sterility control(SC). The plate was incubated at 37°C for 48 to 72 h, after which readings were recorded spectropho-tometrically using an automated plate reader set at 540 nm. The MIC endpoint for AmB was determinedas the lowest drug concentration where there was complete inhibition of growth.

Checkerboard analysis. To determine the drug interaction between AmB and caspofungin, workingsolutions of each drug were prepared in AM3-2% Dex at a concentration four times the targeted finalconcentration in the assay range. One hundred microliters of drug mixture (50 �l of increasingconcentrations of one drug and 50 �l of fixed concentration of the other drug) was mixed and dispensed

Banerjee et al.

Volume 1 Issue 6 e00191-16 msphere.asm.org 12

on February 24, 2021 by guest

http://msphere.asm

.org/D

ownloaded from

in the 96-well microtiter plates. Final concentrations of antifungals for MIC detection ranged from0.0009 �g/ml to 0.5 �g/ml and 0.6 �g/ml to 40 �g/ml for AmB and caspofungin, respectively. Wells weredesignated GC and SC as explained above. One hundred microliters of yeast inoculum (with final cellconcentration of 1 � 105 CFU/ml) was dispensed in each well except the SC, following which trays wereincubated at 37°C for 72 h. Readings were recorded spectrophotometrically. The MIC endpoint for AmBwas determined as explained above, whereas for caspofungin, the MIC was defined as the lowestconcentration of drug tested alone and in combination at which the turbidity in the well was 50 to 80%less than in the control well (GC). The FIC of each drug and the FIC index were calculated from therespective MIC and FIC values to determine the interaction status of the drugs tested (17).

Galleria mellonella killing assay. G. mellonella larvae were allowed to acclimate in the dark at 30°Cand 37°C overnight prior to being injected. Ten larvae of required weight (250 25 mg) were selectedfor each group in the assay. Inoculum was prepared by washing overnight cultures of wild-type, ura1Δmutant, and C-URA1 cells three times and resuspending them in PBS. Cells were counted in a hemocy-tometer slide and diluted with PBS to yield an inoculum size of 1 � 108 cells that was injected by a 10-�lHamilton syringe. Ten-microliter aliquots of the inoculum were injected into the hemocoel of each larvavia the last proleg after disinfecting the injected area with an alcohol swab. One group of larvae wasinjected with PBS to serve as a control to monitor killing due to physical injury. The larvae were incubatedin the dark at 30°C and 37°C and monitored daily to record the number of dead larvae. Death wasconsidered when the caterpillars melanized and/or failed to respond to touch. Killing curves were plottedover time to depict survival of the different strains at the two different temperatures.

Statistical analyses. GraphPad Prism 6.0 software was used for statistical analyses of capsule sizesusing the grouped analyses (two-way analysis of variance [ANOVA]), and standard deviation was plottedon the graph. Kill curves in the G. mellonella infection model experiment were analyzed using theKruskal-Wallis analysis (ANOVA on ranks), and a P value of �0.05 was considered significant. Experimentswere performed in triplicates.

ACKNOWLEDGMENTSWe are thankful to Joseph Heitman, Jennifer K. Lodge, and the Fungal Genetic StockCenter for providing the strains used in this study.

FUNDING INFORMATIONThis work, including the efforts of John C. Panepinto, was funded by HHS | NationalInstitutes of Health (NIH) (1R01 AI089920-01A1).

REFERENCES1. Park BJ, Wannemuehler KA, Marston BJ, Govender N, Pappas PG,

Chiller TM. 2009. Estimation of the current global burden of cryptococ-cal meningitis among persons living with HIV/AIDS. AIDS 23:525–530.http://dx.doi.org/10.1097/QAD.0b013e328322ffac.

2. Perfect JR, Dismukes WE, Dromer F, Goldman DL, Graybill JR, HamillRJ, Harrison TS, Larsen RA, Lortholary O, Nguyen MH, Pappas PG,Powderly WG, Singh N, Sobel JD, Sorrell TC. 2010. Clinical practiceguidelines for the management of cryptococcal disease: 2010 update bythe Infectious Diseases Society of America. Clin Infect Dis 50:291–322.http://dx.doi.org/10.1086/649858.

3. Aoun M. 2000. Standard antifungal therapy in neutropenic patients. IntJ Antimicrob Agents 16:143–145.

4. Saag MS, Graybill RJ, Larsen RA, Pappas PG, Perfect JR, PowderlyWG, Sobel JD, Dismukes WE. 2000. Practice guidelines for the man-agement of cryptococcal disease. Infectious Diseases Society of America.Clin Infect Dis 30:710 –718. http://dx.doi.org/10.1086/313757.

5. Brouwer AE, Rajanuwong A, Chierakul W, Griffin GE, Larsen RA,White NJ, Harrison TS. 2004. Combination antifungal therapies forHIV-associated cryptococcal meningitis: a randomised trial. Lancet 363:1764 –1767. http://dx.doi.org/10.1016/S0140-6736(04)16301-0.

6. Bennett JE, Dismukes WE, Duma RJ, Medoff G, Sande MA, Gallis H,Leonard J, Fields BT, Bradshaw M, Haywood H, McGee ZA, Cate TR,Cobbs CG, Warner JF, Alling DW. 1979. A comparison of amphotericinB alone and combined with flucytosine in the treatment of cryptococcalmeningitis. N Engl J Med 301:126 –131. http://dx.doi.org/10.1056/NEJM197907193010303.

7. Schwarz P, Dromer F, Lortholary O, Dannaoui E. 2003. In vitro inter-action of flucytosine with conventional and new antifungals againstCryptococcus neoformans clinical isolates. Antimicrob Agents Che-mother 47:3361–3364.

8. Morrow CA, Valkov E, Stamp A, Chow EW, Lee IR, Wronski A,Williams SJ, Hill JM, Djordjevic JT, Kappler U, Kobe B, Fraser JA.2012. De novo GTP biosynthesis is critical for virulence of the fungal

pathogen Cryptococcus neoformans. PLoS Pathog 8:e1002957. http://dx.doi.org/10.1371/journal.ppat.1002957.

9. Nagy M, Lacroute F, Thomas D. 1992. Divergent evolution of pyrimi-dine biosynthesis between anaerobic and aerobic yeasts. Proc Natl AcadSci U S A 89:8966 – 8970. http://dx.doi.org/10.1073/pnas.89.19.8966.

10. Liu S, Neidhardt EA, Grossman TH, Ocain T, Clardy J. 2000. Structuresof human dihydroorotate dehydrogenase in complex with antiprolifera-tive agents. Structure 8:25–33. http://dx.doi.org/10.1016/S0969-2126(00)00077-0.

11. Björnberg O, Rowland P, Larsen S, Jensen KF. 1997. Active site ofdihydroorotate dehydrogenase A from Lactococcus lactis investigatedby chemical modification and mutagenesis. Biochemistry 36:16197–16205. http://dx.doi.org/10.1021/bi971628y.

12. Baldwin J, Farajallah AM, Malmquist NA, Rathod PK, Phillips MA.2002. Malarial dihydroorotate dehydrogenase. Substrate and inhibitorspecificity. J Biol Chem 277:41827– 41834. http://dx.doi.org/10.1074/jbc.M206854200.

13. McLean JE, Neidhardt EA, Grossman TH, Hedstrom L. 2001. Multipleinhibitor analysis of the brequinar and leflunomide binding sites onhuman dihydroorotate dehydrogenase. Biochemistry 40:2194 –2200.

14. Pizzorno G, Wiegand RA, Lentz SK, Handschumacher RE. 1992. Bre-quinar potentiates 5-fluorouracil antitumor activity in a murine modelcolon 38 tumor by tissue-specific modulation of uridine nucleotidepools. Cancer Res 52:1660 –1665.

15. Phillips MA, Rathod PK. 2010. Plasmodium dihydroorotatedehydrogenase: a promising target for novel anti-malarial chemother-apy. Infect Disord Drug Targets 10:226 –239. http://dx.doi.org/10.2174/187152610791163336.

16. de Gontijo FA, Pascon RC, Fernandes L, Machado J, Jr, Alspaugh JA,Vallim MA. 2014. The role of the de novo pyrimidine biosyntheticpathway in Cryptococcus neoformans high temperature growth andvirulence. Fungal Genet Biol 70:12–23. http://dx.doi.org/10.1016/j.fgb.2014.06.003.

17. Banerjee D, Burkard L, Panepinto JC. 2014. Inhibition of nucleotide

Cell Integrity Defects Synergize with AmB

Volume 1 Issue 6 e00191-16 msphere.asm.org 13

on February 24, 2021 by guest

http://msphere.asm

.org/D

ownloaded from

biosynthesis potentiates the antifungal activity of amphotericin B. PLoSOne 9:e87246. http://dx.doi.org/10.1371/journal.pone.0087246.

18. Zameitat E, Freymark G, Dietz CD, Löffler M, Bölker M. 2007. Func-tional expression of human dihydroorotate dehydrogenase (DHODH) inpyr4 mutants of Ustilago maydis allows target validation of DHODHinhibitors in vivo. Appl Environ Microbiol 73:3371–3379. http://dx.doi.org/10.1128/AEM.02569-06.

19. Zameitat E, Knecht W, Piskur J, Löffler M. 2004. Two different dihy-droorotate dehydrogenases from yeast Saccharomyces kluyveri. FEBSLett 568:129 –134. http://dx.doi.org/10.1016/j.febslet.2004.05.017.

20. Couto SG, Nonato MC, Costa-Filho AJ. 2008. Defects in vesicle coreinduced by Escherichia coli dihydroorotate dehydrogenase. Biophys J94:1746 –1753. http://dx.doi.org/10.1529/biophysj.107.120055.

21. Baumann P, Mandl-Weber S, Völkl A, Adam C, Bumeder I, Oduncu F,Schmidmaier R. 2009. Dihydroorotate dehydrogenase inhibitorA771726 (leflunomide) induces apoptosis and diminishes proliferationof multiple myeloma cells. Mol Cancer Ther 8:366 –375. http://dx.doi.org/10.1158/1535-7163.MCT-08-0664.

22. Zameitat E, Gojkovic Z, Knecht W, Piskur J, Löffler M. 2006. Biochem-ical characterization of recombinant dihydroorotate dehydrogenasefrom the opportunistic pathogenic yeast Candida albicans. FEBS J 273:3183–3191. http://dx.doi.org/10.1111/j.1742-4658.2006.05327.x.

23. Franzot SP, Salkin IF, Casadevall A. 1999. Cryptococcus neoformansvar. grubii: separate varietal status for Cryptococcus neoformans sero-type A isolates. J Clin Microbiol 37:838 – 840.

24. Doering TL. 2009. How sweet it is! Cell wall biogenesis and polysaccha-ride capsule formation in Cryptococcus neoformans. Annu Rev Microbiol63:223–247. http://dx.doi.org/10.1146/annurev.micro.62.081307.162753.

25. Zaragoza O, Rodrigues ML, De Jesus M, Frases S, Dadachova E,Casadevall A. 2009. The capsule of the fungal pathogen Cryptococcusneoformans. Adv Appl Microbiol 68:133–216. http://dx.doi.org/10.1016/S0065-2164(09)01204-0.

26. Keele DJ, DeLallo VC, Lewis RE, Ernst EJ, Klepser ME. 2001. Evaluationof amphotericin B and flucytosine in combination against Candidaalbicans and Cryptococcus neoformans using time-kill methodology.Diagn Microbiol Infect Dis 41:121–126. http://dx.doi.org/10.1016/S0732-8893(01)00297-8.

27. Klepser ME, Wolfe EJ, Pfaller MA. 1998. Antifungal pharmacodynamiccharacteristics of fluconazole and amphotericin B against Cryptococcusneoformans. J Antimicrob Chemother 41:397– 401. http://dx.doi.org/10.1093/jac/41.3.397.

28. Pappalardo MC, Szeszs MW, Martins MA, Baceti LB, Bonfietti LX,Purisco SU, Baez AA, Melhem MS. 2009. Susceptibility of clinicalisolates of Cryptococcus neoformans to amphotericin B using time-killmethodology. Diagn Microbiol Infect Dis 64:146 –151. http://dx.doi.org/10.1016/j.diagmicrobio.2009.02.007.

29. Kraus PR, Boily MJ, Giles SS, Stajich JE, Allen A, Cox GM, Dietrich FS,Perfect JR, Heitman J. 2004. Identification of Cryptococcus neoformanstemperature-regulated genes with a genomic-DNA microarray. EukaryotCell 3:1249 –1260. http://dx.doi.org/10.1128/EC.3.5.1249-1260.2004.

30. Fang J, Yamaza H, Uchiumi T, Hoshino Y, Masuda K, Hirofuji Y,Wagener FA, Kang D, Nonaka K. 2016. Dihydroorotate dehydrogenasedepletion hampers mitochondrial function and osteogenic differentia-tion in osteoblasts. Eur J Oral Sci 124:241–245. http://dx.doi.org/10.1111/eos.12270.

31. Fang JX, Uchiumi T, Yagi M, Matsumoto S, Amamoto R, Takazaki S,Yamaza H, Nonaka K, Kang DC. 2013. Dihydro-orotate dehydrogenaseis physically associated with the respiratory complex and its loss leads tomitochondrial dysfunction. Biosci Rep 33:e00021. http://dx.doi.org/10.1042/BSR20120097.

32. Garcia-Solache MA, Izquierdo-Garcia D, Smith C, Bergman A, Casa-devall A. 2013. Fungal virulence in a lepidopteran model is an emergentproperty with deterministic features. mBio 4:e00100-13. http://dx.doi.org/10.1128/mBio.00100-13.

33. Castedo M, Hirsch T, Susin SA, Zamzami N, Marchetti P, Macho A,Kroemer G. 1996. Sequential acquisition of mitochondrial and plasmamembrane alterations during early lymphocyte apoptosis. J Immunol157:512–521.

34. Mlcochova P, Pelchen-Matthews A, Marsh M. 2013. Organization andregulation of intracellular plasma membrane-connected HIV-1 assemblycompartments in macrophages. BMC Biol 11:89. http://dx.doi.org/10.1186/1741-7007-11-89.

35. Gerik KJ, Donlin MJ, Soto CE, Banks AM, Banks IR, Maligie MA,Selitrennikoff CP, Lodge JK. 2005. Cell wall integrity is dependent onthe PKC1 signal transduction pathway in Cryptococcus neoformans. MolM i c r o b i o l 5 8 : 3 9 3 – 4 0 8 . h t t p : / / d x . d o i . o r g / 1 0 . 1 1 1 1 / j . 1 3 6 5-2958.2005.04843.x.

36. Banks IR, Specht CA, Donlin MJ, Gerik KJ, Levitz SM, Lodge JK. 2005.A chitin synthase and its regulator protein are critical for chitosanproduction and growth of the fungal pathogen Cryptococcus neofor-mans. Eukaryot Cell 4:1902–1912. http://dx.doi.org/10.1128/EC.4.11.1902-1912.2005.

37. Schwarz P, Janbon G, Dromer F, Lortholary O, Dannaoui E. 2007.Combination of amphotericin B with flucytosine is active in vitro againstflucytosine-resistant isolates of Cryptococcus neoformans. AntimicrobAgents Chemother 51:383–385. http://dx.doi.org/10.1128/AAC.00446-06.

38. Nooney L, Matthews RC, Burnie JP. 2005. Evaluation of Mycograb,amphotericin B, caspofungin, and fluconazole in combination againstCryptococcus neoformans by checkerboard and time-kill methodolo-gies. Diagn Microbiol Infect Dis 51:19 –29. http://dx.doi.org/10.1016/j.diagmicrobio.2004.08.013.

39. White RL, Burgess DS, Manduru M, Bosso JA. 1996. Comparison ofthree different in vitro methods of detecting synergy: time-kill, check-erboard, and E test. Antimicrob Agents Chemother 40:1914 –1918.

40. Odds FC, Brown AJP, Gow NAR. 2003. Antifungal agents: mechanismsof action. Trends Microbiol 11:272–279. http://dx.doi.org/10.1016/S0966-842X(03)00117-3.

41. Joly V, Bolard J, Yeni P. 1992. In vitro models for studying toxicity ofantifungal agents. Antimicrob Agents Chemother 36:1799 –1804. http://dx.doi.org/10.1128/AAC.36.9.1799.

42. Phillips MA, Lotharius J, Marsh K, White J, Dayan A, White KL,Njoroge JW, El Mazouni F, Lao YB, Kokkonda S, Tomchick DR, DengXY, Laird T, Bhatia SN, March S, Ng CL, Fidock DA, Wittlin S,Lafuente-Monasterio M, Benito FJG, Alonso LMS, Martinez MS,Jimenez-Diaz MB, Bazaga SF, Angulo-Barturen I, Haselden JN, Lout-tit J, Cui Y, Sridhar A, Zeeman AM, Kocken C, Sauerwein R, Decher-ing K, Avery VM, Duffy S, Delves M, Sinden R, Ruecker A, WickhamKS, Rochford R, Gahagen J, Iyer L, Riccio E, Mirsalis J, Bathhurst I,Rueckle T, Ding X, Campo B, Leroy D, Rogers MJ. 2015. A long-duration dihydroorotate dehydrogenase inhibitor (DSM265) for preven-tion and treatment of malaria. Sci Transl Med 7:296ra111. http://dx.doi.org/10.1126/scitranslmed.aaa6645.

43. Varma A, Edman JC, Kwon-Chung KJ. 1992. Molecular and geneticanalysis of URA5 transformants of Cryptococcus neoformans. InfectImmun 60:1101–1108.

44. Maeng S, Ko YJ, Kim GB, Jung KW, Floyd A, Heitman J, Bahn YS.2010. Comparative transcriptome analysis reveals novel roles of the Rasand cyclic AMP signaling pathways in environmental stress responseand antifungal drug sensitivity in Cryptococcus neoformans. EukaryotCell 9:360 –378. http://dx.doi.org/10.1128/EC.00309-09.

45. Bahn YS, Kojima K, Cox GM, Heitman J. 2005. Specialization of theHOG pathway and its impact on differentiation and virulence of Cryp-tococcus neoformans. Mol Biol Cell 16:2285–2300. http://dx.doi.org/10.1091/mbc.E04-11-0987.

46. Kim SY, Ko YJ, Jung KW, Strain A, Nielsen K, Bahn YS. 2011. Hrk1plays both Hog1-dependent and -independent roles in controlling stressresponse and antifungal drug resistance in Cryptococcus neoformans.PLoS One 6:e18769. http://dx.doi.org/10.1371/journal.pone.0018769.

47. Panepinto J, Liu L, Ramos J, Zhu X, Valyi-Nagy T, Eksi S, Fu J, JaffeHA, Wickes B, Williamson PR. 2005. The DEAD-box RNA helicase Vad1regulates multiple virulence-associated genes in Cryptococcus neofor-mans. J Clin Invest 115:632– 641. http://dx.doi.org/10.1172/JCI23048.

48. Toffaletti DL, Rude TH, Johnston SA, Durack DT, Perfect JR. 1993.Gene transfer in Cryptococcus neoformans by use of biolistic delivery ofDNA. J Bacteriol 175:1405–1411

49. Edman JC, Kwon-Chung KJ. 1990. Isolation of the URA5 gene fromCryptococcus neoformans var. neoformans and its use as a selectivemarker for transformation. Mol Cell Biol 10:4538 – 4544. http://dx.doi.org/10.1128/MCB.10.9.4538.

50. Biasini M, Bienert S, Waterhouse A, Arnold K, Studer G, Schmidt T,Kiefer F, Gallo Cassarino T, Bertoni M, Bordoli L, Schwede T. 2014.SWISS-MODEL: modelling protein tertiary and quaternary structure us-ing evolutionary information. Nucleic Acids Res 42:W252–W258. http://dx.doi.org/10.1093/nar/gku340.

Banerjee et al.

Volume 1 Issue 6 e00191-16 msphere.asm.org 14

on February 24, 2021 by guest

http://msphere.asm

.org/D

ownloaded from